Characterization of Foci and Breakthrough Sites During Persistent and Long‐Standing Persistent Atrial Fibrillation in Patients: Studies Using High‐Density (510–512 Electrodes) Biatrial Epicardial Mapping
Background We previously demonstrated that persistent and long‐standing persistent atrial fibrillation is maintained by activation emanating from foci and breakthrough sites of different cycle lengths (CLs). The purpose of this study was to characterize the behavior of focal and nonrandom breakthrough activation identified during high‐density mapping of atrial fibrillation in these patients.
Methods and Results During open heart surgery, we recorded activation from both atria simultaneously using 510 to 512 epicardial electrodes along with ECG lead II in 12 patients with persistent and long‐standing persistent atrial fibrillation. For each patient, analysis of 32 consecutive seconds of activation from identified focal (sustained and/or intermittent) and nonrandom breakthrough sites was performed. Multiple foci (sustained and/or intermittent) of different CLs were present in both atria in 11 of 12 patients; 8 foci were sustained, and 22 were intermittent. Temporal CL behavior of sustained foci varied over time (≤20 ms of the mean CL). For intermittent foci, no activation periods were due to a spontaneous pause (18 of 22) or activation of the focus by another wave front (11 of 22). All patients had breakthrough activation. Seven patients had 12 nonrandom breakthrough sites. Periods of no breakthrough activation were caused by a spontaneous pause (6 of 12 patients) or activation from another wave front (4 of 12) or were uncertain (5 of 12). Focal and nonrandom breakthrough activation sometimes produced repetitive “wannabe” (incomplete) reentry in 6 of 12 patients.
Conclusions During persistent and long‐standing persistent atrial fibrillation, sustained foci manifested variable CLs. Spontaneous pauses or activation from other wave fronts explained the intermittency of foci and nonrandom breakthrough. Focal and nonrandom breakthrough activation occasionally produced wannabe reentry.
Over the years, many atrial activation mapping studies in patients with atrial fibrillation (AF) have described the presence of foci suggestive of or thought to act as drivers maintaining AF.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 However, these studies did not further characterize the putative focal driver or drivers. Recently, in patients with persistent and long‐standing persistent (LSP) AF, we demonstrated that atrial activation results from wave fronts emanating from multiple focal (sustained and/or intermittent) and breakthrough sites of different cycle lengths (CLs).11 Atrial activation from these sites during the AF resulted largely in collision and fusion of wave fronts at continuously varying sites, the latter because of the different CLs of the activation wave fronts. Thus, the study supported the likelihood that focal and nonrandom breakthrough sites act as drivers. No reentry was seen. The purpose of this study was to characterize further the sustained and intermittent focal and nonrandom breakthrough activation identified during high‐density mapping of AF in patients with persistent and LSP AF.
The research protocol was approved by the institutional review board at University Hospitals Cleveland Medical Center. Twelve patients with persistent and LSP AF (1 month to 9 years in duration) were studied during open heart surgery (Table 1). All patients gave written informed consent before their surgery.
Using standard techniques described previously,11 atrial epicardial mapping studies were performed during open heart surgery in patients with persistent and LSP AF after the heart was exposed using standard surgical procedures. Studies were performed on a beating heart at normothermia either prior to or during cardiopulmonary bypass. All patients underwent a transesophageal echocardiography study prior to mapping the AF to screen for the presence of a left atrial thrombus. If the latter was found, the patient did not undergo mapping. Three electrode arrays with a total of 510 electrodes (first 4 patients) or 512 electrodes (subsequent 8 patients) covering a total area of 92.85 cm2 were placed on the atrial epicardial surface of both atria for simultaneous recording, as described previously.11 The interelectrode distance between each bipolar electrode pair was 1.2 or 1.5 mm, respectively. Atrial electrograms (AEGs) from 510 to 512 electrodes (255–256 bipolar pairs) along with ECG lead II were simultaneously recorded for 1 to 5 minutes during persistent and LSP AF. Data were digitally recorded and processed with an Active Two system (BioSemi). All AEGs were sampled at 1024 Hz and digitized at 24 bits. Data were transferred in real time and stored on a laptop computer for further analysis (CEPAS; Cuoretech Pty Ltd).
Activation Sequence Analysis
Sequential activation maps of persistent and LSP AF were constructed for a period of 32 consecutive seconds per patient using a custom‐designed algorithm that has been previously described in detail.11, 12 All isochronal lines were determined by activation times from recorded bipolar AEGs. Once the activation sequence maps were constructed, the earliest sites of activation compared with their neighbors were identified, and the morphology of the unipolar AEGs were characterized at these sites. All bipolar AEGs were subjected to CL variation and dominant frequency (DF) analyses to detect mean CL, standard deviation, and DF.13 Data are presented as the mean CL±SD because 1000/DF is equivalent to the mean CL.
As described previously,11 a focus was defined as a site (1) that had the earliest activation compared with its neighboring sites, (2) that manifested a QS morphology in the unipolar AEG, and (3) from which wave fronts emanated. A sustained focus was defined as a focus from which wave fronts emanated continuously during the 32 seconds of analysis. An intermittent focus was defined as a focus that was not continuous during 32 seconds of analysis. A breakthrough activation site was defined as a site (1) that had the earliest activation compared with its neighbors, (2) that manifested a unipolar AEG morphology with an initial r or R wave, and (3) from which wave fronts emanated. Nonrandom breakthrough activation was defined as a breakthrough activation site (1) that recurred at the same site, (2) that had at least 1 episode in which there were ≥3 consecutive breakthrough activations, and (3) from which wave fronts emanated. All other breakthrough activation sites were considered random. Reentry was defined as circus movement with head–tail interaction. “Wannabe” reentry was defined as a wave front that circulated around a functional line of block that wanted to become (ie, “wannabe”) a reentrant circuit but could not complete the revolution because it either collided with a wave front from a focal or breakthrough site or encountered a line of block.12
Characteristics of Focal and Breakthrough Epicardial Atrial Activation
For sustained focal sites, the mean CL and DF of bipolar AEG recordings during consecutive 4‐second segments were measured using CL variability detection and DF analyses. To characterize the periods of time when intermittent focal and nonrandom breakthrough sites ceased to initiate a wave front, as demonstrated in the activation sequence maps, both the previous CL and that noninitiating CL at the sites of focal and/or nonrandom breakthrough were measured and compared. When the noninitiating CL interval was <10% of the site's previous CL, it was characterized as an invading activation wave front. When the CL interval was >10% of the previous CL, it was characterized as a spontaneous pause with activation from another wave front.
Data were presented as the mean±SD using Minitab (Minitab Inc).
During persistent and LSP AF, multiple foci (sustained, range 1–2; intermittent, range 1–3) of different CLs were present in both atria in 11 of 12 patients. Figure 1A shows summary data for the temporal behavior of all sustained and intermittent foci during the 32 seconds of analysis from 11 patients. A total of 8 sustained foci (mean CL 170±19 ms; range 142–200 ms; duration 32 seconds) and 22 intermittent foci (mean CL 176±18 ms; range 143–211 ms) were identified in 11 patients. Temporal CL behavior of all 8 sustained foci varied over time. In consecutive 4‐second segments in 6 of 8 foci, there was a maximum consecutive mean CL change of >10 ms. Regarding the temporal behavior of intermittent foci, the duration of individual episodes of intermittent focal activation was variable. Periods of no focal QS activation were caused by either a spontaneous pause (18 of 22 foci) or early activation of the focal site by wave fronts originating from another focus or a breakthrough site (11 of 22 foci) (Table 2). Eight intermittent foci exhibited both a spontaneous pause and activation from an invading wave front during the 32 seconds of analysis.
Breakthrough activation sites were found in all patients, with only 1 patient demonstrating breakthrough sites. A total of 12 nonrandom breakthrough sites (mean CL 165±17 ms; range 136–187 ms) were identified in 7 patients. The disappearance of nonrandom breakthrough sites was due to a spontaneous pause (6 of 12 patients) or activation of the nonrandom breakthrough site by another wave front (4 of 12 patients) (Table 2). The locations of all foci and nonrandom breakthrough sites are summarized in Figure 1B and Table 2. Focal and nonrandom breakthrough sites were found in both atria but were located predominantly in the left atrium.
No epicardial macro reentry was demonstrated in any patient studied; however, wannabe reentry was present in all patients. In 6 of 12 patients, some episodes were generated by focal or nonrandom breakthrough activation. The longest duration per patient of wannabe reentry generated by focal or nonrandom breakthrough activation ranged from 0.4 to 9.2 seconds.
Temporal Characterization of Focal and Nonrandom Breakthrough Activation
Temporal behavior of sustained foci.
Figure 2 shows the temporal behavior of consecutive mean CLs in 4‐second segments during the 32 seconds of recording for the 8 sustained foci. The mean CL of each consecutive 4‐second segment varied over time. The maximum change in consecutive mean CLs for any site was 20 ms. The average mean CL for a site was calculated from the 8 means making up the 32 seconds of recording. The average change in the mean of the mean CL for all sites was 15.5 ms (range 8–20 ms). Six sustained foci had at least 1 consecutive mean CL change of >10 ms, whereas the other 2 had consecutive mean CL changes of <10 ms. Thus, the sustained foci were not characterized by a constant mean CL but rather by a mean CL that fluctuated, on average, 15.5 ms but never >20 ms.
Figure 3 is a representative example of the temporal behavior of the mean CL of a sustained focus (patient 4, sustained focus 2 in Figure 2). Figure 3A shows 32 seconds of a bipolar recording from this sustained focal site. In the right panel, DFs of the bipolar AEGs are shown for the duration of the displayed AEG recording to the left. The shaded areas in the top bipolar AEG panel show the location and duration of each of the 4‐second recordings seen in the 3 panels below. Figure 3B through 3D show lead II and a bipolar AEG from the sustained focus, along with its corresponding unipolar components (U1 and U2), demonstrating a QS morphology in each unipolar AEG. Figure 3B shows the recording from 4 to 8 seconds (DF 6 Hz, mean CL 167 ms). Figure 3C shows the recording from 12 to 16 seconds (DF 5.5 Hz, mean CL 175 ms). Figure 3D shows the recording from 24 to 28 seconds (DF 5.3 Hz, mean CL 187 ms). Overall, the mean CL of this sustained focus varied over time from 167 to 187 ms.
Temporal behavior of intermittent foci and nonrandom breakthrough sites.
The intermittent disappearance of focal and nonrandom breakthrough activation was caused by either an invading wave front from another site (focal or breakthrough) or a spontaneous pause. Figure 4 is a representative example from patient 6 of the disappearance of an intermittent focus, caused in this case by an invading wave front originating from a sustained focus. The top panel of Figure 4 shows the activation sequence maps of 10 consecutive beats during AF recorded from the Bachmann's bundle electrode array, which covers the left atrial appendage, Bachmann's bundle, and the right atrial appendage (RAA). A 10‐ms isochronal color bar from 0 to 220 ms is shown at right. The bottom panel of Figure 4 shows selected bipolar AEGs (sites a through g) recorded simultaneously during AF from 2 foci (sites b and f) and 5 nearby sites (a, c, d, e, and g). As seen in the top panel, because the CL of sustained focal site b is shorter than that of intermittent focal site f, the wave front from sustained focal site b progressively invades the area previously activated by the site f wave front. The site b wave front eventually prevented initiation of activation from intermittent focal site f for several consecutive activation windows. This is clearly demonstrated in the 10 consecutive activation sequence panels. In activation sequence 1 (Figure 4, top left panel), 1 wave front from sustained focal site b propagates toward the left atrial appendage, where it collides with another wave front. It also propagates toward the RAA, where it collides with a wave front from intermittent focal site f near site d. In activation sequence 2, the wave front from sustained focal site b propagates farther toward the RAA and collides with the wave front from intermittent focal site f near site e. In activation sequence 3, the wave front from sustained focal site b propagates even farther toward the RAA and collides with the wave front from intermittent focal site f between sites e and f. Finally, in activation sequence 4, the wave front emanating from sustained focal site b completely invades the area of intermittent focal site f, thereby preventing initiation of activation from that site. Sustained focal activation from site b then dominates activation throughout activation sequence panels 4 to 6 (Figure 4). During activation sequences 4, 5 and 6, a functional line of block (Figure 4, dashed line) and areas of slow conduction (crowding of isochrones) develop. Then, in activation sequence 7, the wave front from sustained focal site b blocks in an area of previous slow conduction, allowing intermittent focal site f to reappear. In activation sequence 8, the wave front from sustained focal site b blocks, presumably, in fact, almost surely, due to refractoriness from the prior activation from site f. In activation sequence 9, the wave front from sustained focal site b once again collides with the wave front from intermittent focal site f, again near site d. In activation sequence 10, the prior activation pattern begins to repeat itself. In addition, wannabe reentry is generated from focal QS activation during activation sequences 6 to 10 in the RAA (note the pivoting wave front around a function line of block in these activation sequences) (Figure 4).
Selected AEGs are shown in the bottom panel of Figure 4, along with burst symbols and propagation arrows illustrating the above description. There were sustained 170‐ms CLs at focal site b and intermittent 200 ms CLs at focal site f; however, when sustained focal site b invaded intermittent focal site f, the 3 CLs recorded at site f (underlined CLs at the bottom of Figure 4) followed the CLs of sustained focal site b (<10% of previous CL). When focal activation of site f reappeared, it returned to its intrinsic, longer CL.
Figure 5 is a representative example in patient 4 of the disappearance of a focus due to a spontaneous pause. The top panel shows a magnified section of the activation sequence maps for 5 consecutive beats during AF. The bottom panel shows bipolar AEGs at selected sites a, b, and c located in the right atrium showing a spontaneous pause of focal activation at site c, thereby allowing activation of site c by another wave front. The bipolar and both unipolar (U1 and U2) AEGs from focal site c are shown. Activations 1, 2, and 3 show that a wave front from focal site c propagated to site a at a CL of 171 ms. The unipolar AEGs from focal site c demonstrate a QS morphology, indicating that the wave front was emanating from this site. After a spontaneous pause at site c, denoted by a longer CL (>10% of previous CL), activations 3 and 4 show an invading wave front propagating in the opposite direction from site a to site c. The unipolar AEGs from site c now demonstrate an rS morphology (denoted by an asterisk in Figure 5), indicating that the site was passively activated by another wave front from a different direction, as demonstrated in the activation sequence maps (activations 4 and 5).
Spatial Characterization of Focal and Nonrandom Breakthrough Activation
The wave fronts generated from focal and breakthrough sites largely produced collision and merging with other wave fronts at continuously varying sites.11 This nonuniform conduction sometimes produced areas of slow conduction and functional lines of block. Occasionally, as a result, activation from a focal or nonrandom breakthrough site generated a pivoting wave front around a functional line of block, producing wannabe reentry. In Figure 4, when a functional line of block (dashed line) was present in the RAA (activation sequences 6, 7, 9, and 10), wannabe reentry was seen to develop from activation from a focal QS site. Figure 6 shows how a repetitive wannabe reentry pattern developed over a period of ≈10 seconds during activation from a focus in the left atrium (patient 11). The top panel of Figure 6 shows a detailed area of activation sequence maps during 7 AF activations. The first 2 consecutive beats demonstrate the presence of slow conduction, denoted by crowding of isochrones. Then, as the slow conduction increases over the next 2 seconds (not shown), the following 2 consecutive activation maps show further slowing of conduction (denoted by further crowding of isochrones). The latter is associated with the formation of a short functional line of block and with curvature of 2 different wave fronts originating from the same focus, which shortly results in their collision. Then, after 8 seconds, the next consecutive activation maps show 3 consecutive beats of wannabe reentry cycles that occur around a functional line of block that formed in the area of slow conduction.
The bottom panel of Figure 6 shows selected bipolar AEGs recorded simultaneously from a focal site (c) and 5 nearby sites (a, b, d–f) that are in or near an area of slow conduction (activation sequences 1 and 2) and a functional line of block (dashed lines, activation sequence 3–7) during AF. The wave front generated from repetitive activation emanating from focal site c propagated in a radial‐type activation pattern but with different conduction velocities, denoted by different intervals between isochrones. After a functional line of block developed, repetitive wannabe reentrant activation formed. This occurred when propagation from the focus traveling along one side of the functional line of block pivoted around one end but failed to complete the rotation when it collided with the functional line of block. In addition, wannabe reentry sometimes developed from activation from a nonrandom breakthrough site (Figure 7). Focal and nonrandom breakthrough activations generating repetitive wannabe reentry were present in 6 of 12 patients.
Using high‐density simultaneous biatrial epicardial mapping in patients with persistent and LSP AF, we characterized focal (sustained and/or intermittent) and nonrandom breakthrough activation during 32 seconds of classical activation sequence maps in 12 patients. Our results demonstrated in these patients (1) that atrial activation was caused by wave fronts emanating from foci and/or breakthrough sites of different but short CLs, (2) that the mean CL of sustained foci varied over time, (3) that the disappearance of both intermittent focal and nonrandom breakthrough activation was caused by either a spontaneous pause or early activation of the site by another wave front, and (4) that focal and breakthrough activation occasionally generated wannabe reentry.
Previous Mapping Studies During Persistent and LSP AF in Patients
Using endocardial catheter mapping in paroxysmal and/or persistent AF, sustained focal sources have been identified by QS morphology activation9 or periodically organized activation.7, 10 In addition, sustained and intermittent focal sources were found in patients with paroxysmal and persistent AF.5 In several epicardial mapping studies in patients with persistent and LSP AF, focal activation was identified in the left and/or right atrium,1, 2, 3, 4 and repetitive activation patterns were found in the left atrium,14, 15, 16, 17 suggestive of a driver mechanism. In addition, high‐density sequential area epicardial mapping18, 19 in patients with LSP AF found random and intermittent focal wave fronts, but no sustained focal activation was found. A recent endoepicardial right atrial mapping study in patients with AF found that intermittent focal wave fronts (35% focal and 65% breakthrough) frequently appeared in the endocardium and epicardium.20 A noninvasive mapping study6 showed that in paroxysmal and persistent AF patients, simultaneous intermittent focal activation from areas near the pulmonary veins (69% of patients) and from nonpulmonary vein areas (62% of patients) was present. Recently, in a noninvasive mapping study in patients with persistent and LSP AF,8 intermittent QS focal activation was identified in both atria.
Our finding of the presence of sustained and intermittent foci and breakthrough activation during AF concurs, in whole or in part, with the findings of others; however, most previous studies recorded only bipolar AEGs, recorded for short periods, and were not well characterized. Moreover, the recording resolution of most studies was modest at best. Without the determination of unipolar QS morphology, we can suggest only that repetitive focal wave fronts identified in other studies could correspond to our finding of atrial activation from focal and/or nonrandom breakthrough sites. Furthermore, wannabe reentry may explain why some low‐resolution endocardial mapping data have been interpreted as showing reentry and why ablation of the putative “rotor” core may be efficacious, as perhaps focal and/or breakthrough sites were actually ablated. Finally, the focal and breakthrough‐initiated wannabe reentry that we described has some similarity to a noninvasive mapping study8 in which focal activation initiated reentry.
The efficacy of both catheter and surgical ablation of persistent and LSP AF has been suboptimal, mainly because of the complexity of atrial activation during AF and the absence of clear targets. Understanding the mechanism presents opportunities for new approaches to the treatment of persistent and LSP AF, especially for ablation. Study of the mechanisms maintaining AF has been limited by the difficulty in mapping this complex arrhythmia and by questions about recording techniques and analysis algorithms used.21, 22, 23, 24, 25, 26 Our data describe a new paradigm for the mechanism that maintains persistent and LSP AF, to which we have added the temporal and spatial behavior of the focal and nonrandom breakthrough sites from which the wave fronts emanate to produce AF. Finally, the data from our study offer the potential for improving current ablation approaches to treating persistent and LSP AF by targeted ablation of focal and nonrandom breakthrough sites.
AEGs from the endocardium and the atrial septum were not obtained, precluding identification and characterization of the source of the epicardial breakthroughs and perhaps even the focal QS activations. We did not record from the pulmonary veins, but our left atrium and Bachmann's bundle recording arrays basically surround the pulmonary veins. A unipolar QS electrogram might not identify the exact location of a focus because the impulse may originate from an area close by from which conduction does not generate enough of a signal from the atrial myocardium to be appreciated in the unipolar AEG. The same may also apply to activation from a subepicardial reentrant circuit. Although we characterized some foci as continuous over the 32 seconds of analysis, we do not know if, ultimately, they became intermittent. We cannot rule out the possibility that breakthrough activation was the result of reentry, including subepicardial reentry and/or foci originating at a distance from the breakthrough site.
In our high‐density, simultaneous, biatrial, epicardial mapping study in patients with persistent and LSP AF, each sustained focus manifested minimal fluctuations in CLs over time. Intermittent foci and nonrandom breakthrough sites manifested variability in periods of activity. Spontaneous pauses or activation from other wave fronts explained the intermittency of the intermittent foci and nonrandom breakthrough. Focal and breakthrough activation sometimes generated wannabe reentry.
Sources of Funding
This work was supported in part by grants from R01 HL074189 from the National Institutes of Health, National Heart, Lung, and Blood Institute, Bethesda, MD; BRTT/WCI TECH 05‐066 from the Ohio Wright Center of Innovations, a Third Frontier program from the State of Ohio, Columbus, OH; and from the Jennie Zoline, Blue Dot, Glenstone, Frank & Gerry Pearl, and MCJ Amelior Foundations.
- ↵Holm M, Johansson R, Brandt J, Luhrs C, Olsson SB. Epicardial right atrial free wall mapping in chronic atrial fibrillation. Documentation of repetitive activation with a focal spread—a hitherto unrecognised phenomenon in man. Eur Heart J. 1997;18:290–310.
- ↵Takahashi Y, Hocini M, O'Neill MD, Sanders P, Rotter M, Rostock T, Jonsson A, Sacher F, Clémenty J, Jaïs P, Haïssaguerre M. Sites of focal atrial activity characterized by endocardial mapping during atrial fibrillation. J Am Coll Cardiol. 2006;47:2005–2012.
- ↵Cuculich PS, Wang Y, Lindsay BD, Faddis MN, Schuessler RB, Damiano RJ Jr., Li L, Rudy Y. Noninvasive characterization of epicardial activation in humans with diverse atrial fibrillation patterns. Circulation. 2010;122:1364–1372.
- ↵Narayan SM, Krummen DE, Clopton P, Shivkumar K, Miller JM. Direct or coincidental elimination of stable rotors or focal sources may explain successful atrial fibrillation ablationon‐treatment analysis of the CONFIRM trial (conventional ablation for AF with or without focal impulse and rotor modulation). J Am Coll Cardiol. 2013;62:138–147.
- ↵Haissaguerre M, Hocini M, Denis A, Shah AJ, Komatsu Y, Yamashita S, Daly M, Amraoui S, Zellerhoff S, Picat M‐Q, Quotb A, Jesel L, Lim H, Ploux S, Bordachar P, Attuel G, Meillet V, Ritter P, Derval N, Sacher F, Bernus O, Cochet H, Jais P, Dubois R. Driver domains in persistent atrial fibrillation. Circulation. 2014;130:530–538.
- ↵Gizurarson S, Dalvi R, Das M, Ha ACT, Suszko A, Chauhan VS. Hierarchical schema for identifying focal electrical sources during human atrial fibrillation implications for catheter‐based atrial substrate ablation. JACC Clin Electrophysiol. 2016;2:656–666.
- ↵Jarman JW, Wong T, Kojodjojo P, Spohr H, Davies JE, Roughton M, Francis DP, Kanagaratnam P, O'Neill MD, Markides V, Davies DW, Peters NS. Organizational index mapping to identify focal sources during persistent atrial fibrillation. J Cardiovasc Electrophysiol. 2014;25:355–363.
- ↵Lee S, Sahadevan J, Khrestian CM, Cakulev I, Markowitz A, Waldo AL. Simultaneous biatrial high‐density (510–512 electrodes) epicardial mapping of persistent and long‐standing persistent atrial fibrillation in patients: new insights into the mechanism of its maintenance. Circulation. 2015;132:2108–2117.
- ↵Sanders P, Berenfeld O, Hocini M, Jais P, Vaidyanathan R, Hsu L‐F, Garrigue S, Takahashi Y, Rotter M, Sacher F, Scavee C, Ploutz‐Snyder R, Jalife J, Haissaguerre M. Spectral analysis identifies sites of high‐frequency activity maintaining atrial fibrillation in humans. Circulation. 2005;112:789–797.
- ↵Sahadevan J, Ryu K, Peltz L, Khrestian CM, Stewart RW, Markowitz AH, Waldo AL. Epicardial mapping of chronic atrial fibrillation in patients: preliminary observations. Circulation. 2004;110:3293–3299.
- ↵de Groot NM, Houben RP, Smeets JL, Boersma E, Schotten U, Schalij MJ, Crijns H, Allessie MA. Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough. Circulation. 2010;122:1674–1682.
- ↵Lee G, Kumar S, Teh A, Madry A, Spence S, Larobina M, Goldblatt J, Brown R, Atkinson V, Moten S, Morton JB, Sanders P, Kistler PM, Kalman JM. Epicardial wave mapping in human long‐lasting persistent atrial fibrillation: transient rotational circuits, complex wavefronts, and disorganized activity. Eur Heart J. 2014;35:86–97.
- ↵de Groot N, van der Does L, Yaksh A, Lanters E, Teuwen C, Knops P, van de Woestijne P, Bekkers J, Kik C, Bogers A, Allessie M. Direct proof of endo‐epicardial asynchrony of the atrial wall during atrial fibrillation in humans. Circ Arrhythm Electrophysiol. 2016;9:e003660.
- ↵Benharash P, Buch E, Frank P, Share M, Tung R, Shivkumar K, Mandapati R. Quantitative analysis of localized sources identified by focal impulse and rotor modulation mapping in atrial fibrillation. Circ Arrhythm Electrophysiol. 2015;8:554–561.
- ↵Buch E, Share M, Tung R, Benharash P, Sharma P, Koneru J, Mandapati R, Ellenbogen KA, Shivkumar K. Long‐term clinical outcomes of focal impulse and rotor modulation for treatment of atrial fibrillation: a multicenter experience. Heart Rhythm. 2015;13:636–641.
- ↵Gianni C, Mohanty S, Di Biase L, Metz T, Trivedi C, Gökoğlan Y, Güneş MF, Bai R, Al‐Ahmad A, Burkhardt JD, Gallinghouse GJ, Horton RP, Hranitzky PM, Sanchez JE, Halbfaß P, Müller P, Schade A, Deneke T, Tomassoni GF, Natale A. Acute and early outcomes of focal impulse and rotor modulation (FIRM)‐guided rotors‐only ablation in patients with nonparoxysmal atrial fibrillation. Heart Rhythm. 2016;13:830–835.
- ↵Kuklik P, Zeemering S, van Hunnik A, Maesen B, Pison L, Lau D, Maessen J, Podziemski P, Meyer C, Schaffer B, Crijns H, Willems S, Schotten U. Identification of rotors during human atrial fibrillation using contact mapping and phase singularity detection: technical considerations. IEEE Trans Biomed Eng. 2017;64:310–318.
- ↵Vijayakumar R, Vasireddi SK, Cuculich PS, Faddis MN, Rudy Y. Methodology considerations in phase mapping of human cardiac arrhythmias. Circ Arrhythm Electrophysiol. 2016;9:e004409.